Deadly Triad Example Check Lite

Problem

Count how many conditions of the deadly triad are present: function approximation, bootstrapping, and off-policy learning. The instability risk is highest when all three are true.

Hint

Tabular on-policy methods avoid the triad entirely.

Test cases

case 1

in: [true,true,true]

expected: 3

case 2

in: [true,false,true]

expected: 2

case 3

in: [false,false,false]

expected: 0

case 4

in: [true,true,false]

expected: 2

Starter codePython

def deadly_triad_count(function_approx, bootstrapping, off_policy):
    # Your code here
    pass

4 test cases · run your code against them in the editor